Wind farm visual impact assessment – part 1

Warning

The viewshed maps on this page are very detailed and may take quite a while to display in many web browsers, and may not render at all on small devices such as smartphones or tablets. They are provided purely for development purposes, and final versions will be optimised for faster display.

Introduction

The ability to assess the visual impact of proposed wind farm developments is important for several reasons:

• current wind turbines are very large structures which are potentially visible over large distances;
• they are moving industrial structures in, typically, otherwise rural settings devoid of other large man-made structures – their movement tends to attract the eye;
• if measures of visual impact can calculated for different proposed configurations of wind turbines (location, number and size of each turbine), then planning decisions can be better optimised;
• wind farms tend to be controversial developments, which can create considerable conflict in local communities and which may be subject to lobbying campaigns, both for and against such developments, which do not use principled or reliable information which has been created using documented, verifiable and reproducible scientific processes.

A typical visual impact assessment factsheet to inform the local community, this one for the NSW Liverpool Ranges wind farm (click image to open the document):

Liverpool Range Wind Farm Visual Impact Assessment Factsheet

Usually proponents (developers) of wind farms commission specialist agencies to provide such visual impact assessments on their behalf, both to inform the local community and other stakeholders, and for inclusion in the EIS (environmental impact statement) and associated documentation required by planning processes. These visual impact assessments are typically use methodologies described in relevant government guidlines for such assessments, such as the 2016 NSW Department of Planning and Environment Wind Energy: Visual Assessment Bulletin.

Visual impact assessment is a complex, multi-disciplinary subject, involving physics, earth sciences, psychology and psychometrics, and social factors. These series of blog posts are not intended to be a comprehensive introduction or treatment of the subject – rather, they merely document the development of some visual impact assessment capabilities by members of Oberon Citizen Science Network, which can be used to provide information about the potential visual impact of proposed wind farms in teh Oberon region.

Note

The examples provided in this series of blog posts use the locations of 47 wind turbines in the proposed Palings Yard wind farm. However, the size of each turbine has been modelled on the turbine type used in a scientific paper by James F. Palmer published in the peer-reviewed journal Landscape and Urban Planning. We will be referring to that paper repeatedly. However, the actual parameters for specific wind farm proposals can easily be substituted and the calculations re-run with minimal effort, and we will do that in due course. Visual impact assessment methods described in other peer-reviewd papers will also be implemented in due course to add to OCSN’s capacity in this respect.

Viewsheds

The most basic type of visual impact assessment is the calculation of the viewshed of a structure. The viewshed is all the points in the surrounding landscape which can be seem from the top (or any point on) a structure such as a building. tower or a wind turbine. If the the structure is a tower used for transmitting radio waves or an obseravtion tower (such as a fire observation tower), then what can be seen from the tower by direct line-of-sight is of obvious interest.

However, we are usually not very interested in what can be seen from, say, the nacelle (hub) of a wind turbine, let alone from the tips of its blades. But we are interested in from which points in the landscape the blade tips, or the nacelle, or indeed the entire span of the wind turbine blades can be seen. This is called the inverse viewshed or reverse viewshed. In fact, it is symmetrical – if a point in the landscape can be seen from a the top of a tower, then the top of that tower can also be seen from that point in the landscape.

Calculating the viewshed

Viewshed calculations rely on digital elevation maps (DEMs). Theses are maps of the elevation of the Earth’s surface, usually created from satellite LiDAR (laser radar) sensor data. The elevation (that is, height above sea level) data they contain is accurate to around +/- 1m or less, and they are available for various spatial resultions from 1m by 1m patches up. For all the calculations here, we are using the 5m by 5m resolution DEMs provided by the NSW government, freely available from the NSW Spatial Information web portal. DEM files for Oberon and the eight regions surrounding it were downloaded and used in these blog posts. These DEMS were then stitched together to form a larger DEM, comprising the follow smaller DEMs: Wallerawang, Orange, Bathurst, Blayney, Oberon, Crookwell, Katoomba, Taralga, Burragorang.

The resulting DEM, viewed as a two-dimensional plot, looks like this:

Here is part of the Oberon-and-environs DEM visualised in three dimensions, with some hypothetical wind turbines shown as icons (but to-scale, these are the actual size of 240m tall turbines) in the Palings Yard wind farm area.

Line-of-sight calculations

Using these DEMs, a viewshed from any point on the map can be calculated by determining whether the intervening terrain will obstruct the view from the view point to any other point on the map. The height of the view point can be specified and added to the terrain elevation at its location. Similarly, the points being viewed in the landscape can be set at a height above the terrain elevation at each point. Since we are interested in reverse viewsheds, we set the observation elevation at 1.5m above the terrain elevation, since that is an average eye level.

A simple illustration of line-of-sight viewshed determination

Of course, this is a simplified model because it only uses bare-earth elevations – that is, the heights of vegetation or intervening buildings are not included in the calculations. As such, the viewsheds calcualted are “worst-case”, and trees etc may reduce the actual viewsheds. However, NSW government guidelines state that bare-earth elevations should be used in visual impact assessments, precisely because they are worst-case. In fact, adding in land-cover elevation information dramatically complcates the computations required, for reasons we won’t cover here. As it is, the viewshed computations involve billions of calculations. Fortunately, efficient algorithms have been developed to make such computations feasible even on laptop computers, although they can still take many hours to complete.

The Earth is not flat…

Of course, reality is more complicated than what is described above. Apart from trees and other ground cover, two other factors need to be taken into account when calculating viewshed: the curvature of th eEarth, and refraction of light by the atmosphere. The Earth curvature issue is intuitively obvious – if we look twards the horizon, the surface of the Earth falls away and thus the horizon is a finite distance away – in fcat, on a clear day you cannot see forever. Another issue is that due to the Earth’s curvature, as we look towards the horizon (that is, at approximately a tangent to the Earth’s surface), our gaze transects increasingly higher altitudes, with thinner and thinner air. This reduction in air density create a refraction lens effect which partially counteracts the effect of the curvature of the Earth on a viewshed, as shown in the diagram below.

Earth curvature and atmospheric refraction effects on line-of-sight viewsheds

The viewshed calculation software we have used takes both of these factors into account using a simple but quite adequate correction mechanism. So let’s look at a viewshed.

Illustrating the effect of Earth curvature and atmospheric refraction correction in viewshed calculations

The map below shows the calculated viewshed for just one of the wind turbines proposed for the Palings Yard wind farm, south of Oberon. The viewshed is for a blade tip at 170.5m elevation.

By toggling on and off the two viewshed layers on the map below, you can compare the results with the Earth curvature and atmospheric refraction correction disabled, and with itenabled using the widely-accepted compensationco-efficient of 0.85714.

It is easy to see that without the correction, the calculated viewshed is much larger because the curvature of the Earth is ignored (although refraction by the atmosphere counteracts the effects of curvature, but only partially).

Viewshed shaded by distance from turbine

Although a simple binary (all-or-nothing) viewshed is useful, it fails to take into account distance. Obviously, the further away you are from an object, the smaller it appears. Also, atmospheric conditions such as haze from smoke or dust (or pollen) can reduce visibility of objects at larger distances (and fog will reduce visibility at even short distances). However, atmospheric conditions are variable and we don’t take them into account in our clculations – we assume a “worst case” of perfectly clear air. But the effcet of distance can be taken into account

Below is a map of the viewshed as above (with curvature and refraction corrections enabled), but shaded by distance from the hypothetical wind turbine, to de-emphasise distant viewpoints. This is a first step towards better represention of visual impact.

Viewshed shaded by distance bands from turbine

The continous distance shading shown above is pretty, but it may be useful to categorise distance into bands. This is typically done in visual impact assessments.

Below is the same viewshed as above (with curvature and refraction corrections), but this time shaded by distance categorised into bands as suggested by James F. Palmer in the paper mentioned above – howeverfor illustrative purposes we have extended the bands to 64.4 km (the bands suggested by Palmer stop at 32.2km, which is probably reasonable)

Viewshed distance bands as separate layers

Finally, we can create a variation of he map above in which each distance zone is represented as a separate layer which can be toggled on and off. This may not be particularly useful.

Next steps

In the next in this series of blog posts we will look at the calculation of compound viewsheds – that is, the viewshed for a collection of wind turbines, not just a single one.

Liverpool Range Wind Farm Visual Impact Assessment Factsheet A simple illustration of line-of-sight viewshed determination Earth curvature and atmospheric refraction effects on line-of-sight viewsheds